<jats:title>Abstract</jats:title><jats:p>Genome mining is a powerful method for finding novel secondary metabolites. In our study on the biosynthetic gene cluster for the cyclic octapeptides surugamides A–E (inhibitors of cathepsin B), we found a putative gene cluster consisting of four successive non‐ribosomal peptide synthetase (NRPS) genes, <jats:italic>surA</jats:italic>, <jats:italic>surB</jats:italic>, <jats:italic>surC</jats:italic>, and <jats:italic>surD</jats:italic>. Prediction of amino acid sequence based on the NRPSs and gene inactivation revealed that surugamides A–E are produced by two NRPS genes, <jats:italic>surA</jats:italic> and <jats:italic>surD</jats:italic>, which were separated by two NRPS genes, <jats:italic>surB</jats:italic> and <jats:italic>surC</jats:italic>. The latter genes are responsible for the biosynthesis of an unrelated peptide, surugamide F. The pattern of intercalation observed in the <jats:italic>sur</jats:italic> genes is unprecedented. The structure of surugamide F, a linear decapeptide containing one 3‐amino‐2‐methylpropionic acid (AMPA) residue, was determined by spectroscopic methods and was confirmed by solid‐phase peptide synthesis.